Omni-MCP
A universal resource server that enables AI models to mount local folders, scrape web pages into Markdown, and query SQLite databases without writing code. It features a real-time TUI dashboard and persistent configuration for managing diverse knowledge sources.
README
<div align="center">
⚡ Omni-MCP
🔌 The Universal MCP Resource Server
Mount any folder, URL, or database. Feed any AI. Zero code required.
</div>
🚀 Highlights
- 📂 Folder Mounting — Mount any directory with one command;
.txt/.mdfiles become instantly readable - 🌐 Web Scraping — Provide a URL, auto-convert to Markdown, cached and persisted
- 🗄️ SQLite Adapter — Mount a database, auto-expose its schema, safely run SELECT queries
- 📟 TUI Dashboard — Ink-powered terminal UI with real-time status, mounts, and AI activity
- 💾 Persistent Config — Mount once, automatically restored on restart
🛠️ Quick Start
1. Install
git clone https://github.com/DEM-YU/omni-mcp.git
cd omni-mcp
npm install
2. Connect to Your AI Client
<details> <summary>🌌 <b>Antigravity</b> (<code>~/.gemini/antigravity/mcp_config.json</code>)</summary>
{
"mcpServers": {
"omni-mcp": {
"command": "npx",
"args": ["tsx", "/path/to/omni-mcp/src/index.ts"],
"cwd": "/path/to/omni-mcp"
}
}
}
</details>
<details> <summary>🧠 <b>Claude Desktop</b> (<code>claude_desktop_config.json</code>)</summary>
{
"mcpServers": {
"omni-mcp": {
"command": "npx",
"args": ["tsx", "/path/to/omni-mcp/src/index.ts"],
"cwd": "/path/to/omni-mcp"
}
}
}
</details>
3. Start Talking
Once connected, just tell your AI:
"Mount the folder /Users/me/research as a knowledge base"
"Scrape this page: https://example.com"
"Mount the database /path/to/data.db and query all users"
📂 Supported Sources
Local Folders
| Tool | Description |
|---|---|
mount_folder |
Mount a local folder, exposing .txt / .md files |
add_new_source |
Alias for mount_folder |
unmount_folder |
Unmount a previously mounted folder |
Web Pages
| Tool | Description |
|---|---|
mount_url |
Fetch a web page, convert to Markdown, cache and expose as a resource |
SQLite Databases
| Tool | Description |
|---|---|
mount_sqlite |
Mount a SQLite database (read-only), auto-expose schema |
query_sqlite |
Safely execute SELECT queries, returns JSON (max 100 rows) |
General
| Tool | Description |
|---|---|
list_mounts |
List all mounted folders, web pages, and databases |
📟 TUI Dashboard
A live terminal dashboard renders on stderr at startup:
- Status Indicator — Server online / starting
- 📂 Folders — Mounted folder tree
- 🌐 Web Pages — Mounted URL list
- 🗄️ Databases — Connected database list
- ⚡ Live Activity — Flashes in real-time when AI reads a resource
💡 TUI renders to
stderr; MCP JSON-RPC flows throughstdout/stdin— zero interference.
📁 Project Structure
omni-mcp/
├── src/
│ ├── index.ts # MCP Server + all tool handlers
│ ├── events.ts # Event bus (Server ↔ Dashboard)
│ └── dashboard.tsx # Ink TUI dashboard
├── test-resources/ # Sample files for testing
├── config.json # Auto-generated persistence (gitignored)
├── package.json
├── tsconfig.json
└── README.md
📄 License
MIT © Brooks
<div align="center">
Built with 🧡 for the AI-native developer workflow.
If Omni-MCP saved you from copy-pasting, give it a ⭐
</div>
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。